Grain boundary segregation prediction with a dual-solute model

Solute segregation along grain boundaries (GBs) profoundly affects their thermodynamic and kinetic behaviors in polycrystalline materials. Recently, the spectral approach has emerged as a powerful tool to predict GB segregation. However, previous GB segregation predictions using this method relied heavily on the single-solute (SS) segregation energy spectrum without solute-solute interactions, which were often incorporated through a fitting parameter. In this paper, we develop a dual-solute (DS) model whose segregation energy spectrum intrinsically incorporates solute-solute interactions. It was first validated for GB segregation prediction in the Al-Mg system and then extended to several other binary alloy systems. The DS model shows significant improvement over the SS model and can accurately predict the real segregation states obtained by hybrid molecular dynamics/Monte Carlo simulations within a broad temperature range with different solute concentrations before forming secondary phases. This DS model provides an effective method for accurately predicting GB segregation in nanocrystalline metals.